 |
Description  |
|
|
FIELD OF THE INVENTION
The present invention relates to an image area discriminating system, which
discriminates a character area from a halftone area on an original
containing a character image and a halftone image and improves the
reproductivity of each area through the area discrimination.
BACKGROUND OF THE INVENTION
FIG. 17 is a block diagram showing an arrangement of a digital color image
processing system. FIG. 18 is a block diagram showing an arrangement of a
conventional edge processing circuit. FIGS. 19(a)-(c) are block diagrams
showing an arrangement of a hue detect circuit and related tables. FIGS.
20(a) through (c) are a graph and explanatory diagrams, which are useful
in explaining a character spread phenomenon. FIGS. 21(a) through (c) are
explanatory diagrams for explaining edge emphasis processing.
Generally, a color copying machine exercises a developing process of Y
(yellow), M (magenta), C (cyan), and K (black), to reproduce a full color
image of a color original. To store full color image data gathered by a
single scan of a color image of an original, a considerably large memory
capacity is required. To avoid this, in the conventional developing
process, the machine scans the color original separately for each color,
and executes signal processing.
In image reading, a line sensor optically reads an image to gather image
data in terms of color separated signals of B (blue), G (green), and R
(red). The separated color signals, as shown in FIG. 17, pass through an
END converter 501 and a color masking (color correction) 502, and are
transformed into color toner signals Y, M and C. Then, the toner signals
enter a UCR 503. In the UCR, the black (K) generation and the under color
removal are carried out. The toner signal as generally designated by X
passes through a hue separation type non-linear filter section, TRC (tone
adjustment) 510, and SG (screen generator) 511, and is converted into
binary data. The binary signal is used to control a laser beam that expose
a photosensitive member. The images of the respective colors are
superposed by the mesh-dot gradation, to reproduce the full color image.
In the images handled by a digital color image processing system, a binary
image, such as characters and lines, and a halftone image, such as
photographs and mesh-dot printing materials, usually coexist. To obtain a
binary image of high sharpness, the original image containing such
different images is subjected to edge emphasis processing, which is based
on non-linear filter processing. As regards the edge emphasis processing,
there have been many proposals. One of those proposals is the arrangement
of FIG. 17, which is provided with a hue separation type non-linear filter
section.
The filter section, as shown, receives the image data signal X of a
developing color as selected from Y, M, C, and signals according to the
developing process. The toner signals are generated through black
generation and under color removal processing. The image data signal X is
branched into two routes. The data signal X flowing through one of the
routes enters a smoothing filter 504 where it is smoothed. The data signal
X is also edge emphasized by the combination of a "r (gamma)" conversion
506, an edge detect filter 507, and an edge emphasizing LUT 508. The data
signals from the two routes are added together by an adder 509, which in
turn produces a non-linear filter signal. An arrangement of the edge
emphasis processing circuit is shown in FIG. 18.
In edge processing, the hue detect circuit 505 detects the hue of an input
image, and determines whether the developing color at that time is a
necessary color or an unnecessary color. If the input image is a black
area, the chromatic signals of Y, M, and C are not edge emphasized, but
only the color signal of K is emphasized according to an edge quantity.
As shown in FIG. 19(a), the hue detect circuit 505 is made up of a
max./min. circuit 512 for obtaining the maximum and minimum values of the
toner signals Y, M, and C, a multiplexer 513 for selecting a developing
color, a subtractor 514 for calculating a difference between the maximum
and minimum values, another subtractor 515 for calculating a difference
between the minimum value and a developing color, and comparators 516 to
518, which compare input signals with threshold values. When the input
signals are larger than the threshold values, the comparators produce
signals r, m, c', m', an y' with logic value "1".
The hue detect circuit recognizes a hue by using a hue decision table as
shown in FIG. 19(b). Further, it determines whether the developing color
is a necessary color of logic "1" or an unnecessary color of logic "0" by
using a necessary/unnecessary color decision table shown in FIG. 19(c).
The hues that are output as the result of the hue determination, are eight
colors, (white), Y, M, C, B, G, R, and K, that are used as normal
character colors.
As seen from the hue decision table, if the hue is B, the necessary
developing colors "m" and "c", and the remaining developing colors are
unnecessary. In this case, during a necessary color cycle, the edges of
the signal are emphasized by the LUT (1) of the edge emphasis LUT 508.
During an unnecessary color cycle, the edges are not emphasized by the LUT
(2) of the LUT 508.
As described above, in edge emphasis processing, the hue of the input
signal is discriminated by comparing the input signal with the threshold
value "th." Depending on the comparison result, an edge detect signal is
converted, by the edge emphasis LUT, into an edge emphasis signal.
Meanwhile, the MTF (modulation transfer function) characteristic of the
IIT (image input terminal) becomes poor as frequency becomes high, as
shown in FIG. 20(a). The degree of degradation of the MTF also changes
depending on the color and the main and vertical scan directions. When the
MTF is degraded, an optical density distribution curve on "a" an original
is flattened to be a curve "b" (see FIG. 20(b)). In detecting a hue, the
signal "b" is compared with the threshold value "th," and the hue is
determined on the basis of the comparison result. Accordingly, the signal
whose hue is recognized has a width w', which is much wider than the width
"w" of the original signal. This defines a range of the edge emphasis
processing. On the basis of the determination result, an edge emphasis
signal "d" as shown in FIG. 20(c) is added to it, to emphasize the edges.
Consequently, it is reproduced in the form of a widened character as
indicated by "c" in FIG. 20(b). The character widening is caused not only
by the IIT, but also by developing material, developing method, developing
characteristic and the like.
When compared to the conventional edge emphasis system in which the color
signals of Y, M, C, and K are all subjected to the edge emphasis
processing, the edge emphasis system as mentioned above improves the
reproduction quality of a black character, but the smoothing signals are
left in the Y, M, and C signals. As indicated by the edge emphasis LUT 508
shown in FIG. 18, the necessary color is emphasized by the LUT (1), while
the unnecessary color is removed by the LUT (2). Accordingly, an edge
emphasis processing signal is generated that does not emphasize the colors
of Y, M, and C of a filter input signal of a black character (as shown in
FIG. 21(a)) does emphasize only the black signal K. In the smoothing
filter, a smoothing processing signal resulting from smoothing all of the
color signals Y, M, C, and K is generated, as shown in FIG. 21(b). When
finally composed the smoothing signal of Y, M, C, and K is as shown in
FIG. 21(c).
Usually, even in the case of the black character, the signal contains not
only the K signal but also the Y, M, and C signals. The smoothed colors of
Y, M, and C appear at the edge portions. Thus, the black character cannot
be reproduced by a single color of K. In connection with the case of the
single color reproduction, the instant case suffers from color change and
a loss of color purity, which are due to widening of lines, impaired
registration, and the like. The resultant image will not be sharp.
In case where there are originals containing binary images, such as
characters and lines, and halftone images, such as photographs and
mesh-dot printing materials, and the type of the image can be designated
for each original or each area, it is possible to select optimum
parameters for the respective types of images. In the case of an image of
the type in which the binary image and the halftone image coexist (this
type of the image will be referred to as an integrated image original),
the parameters selected are those allowing both types of images to be
reproduced. Accordingly, the binary image and the halftone image cannot be
individually processed in the best conditions, and hence satisfactory
images are hard to obtain. In the case of the binary image, the edge
emphasis is weak, and the sharpness of the characters is lost. In the case
of a black character, the edge portions and small characters are blurred.
In the case of the halftone image, the frequencies near the edge detect
frequency ar emphasized. This impairs the smoothness of the halftone
image, and causes unpleasant Moire to appear in the image. Additionally
the edges are unnaturally emphasized. Thus, the resultant image looks hard
and rough.
SUMMARY OF THE INVENTION
Accordingly, an object of the present invention is to improve the precision
for discriminating areas in an integrated image original.
Another object of the present invention is to enable the area
discrimination to be corrected for every block of the image.
A further object of the present invention is to enable both a character
image and a halftone image to be reproduced with high image quality.
An additional object of the present invention is to make it easy to
discriminate image areas within a block of the image. These and other
objects are obtained by an area discriminating system for use in an image
processing system capable of processing an image signal including
character images signals and halftone images signals comprising hue
determining means for determining the hues of the images represented by
the image signal and for producing hue present signals for each color of a
selected number of colors which is a component of the hues of the images
and a hue absent signals for each color of the selected number of colors
which is not a component of the hues of the images, edge detecting means
for detecting edge portions of images represented by the image signal and
for producing edge signals having values representing the edge portions,
and edge emphasis means receiving the hue present signals, the hue absent
signals, and the edge signals and for producing edge emphasized signals
for each hue included in the portion of the image represented by said edge
signals. An image determining means determines whether the image
represented by the image signal is a halftone image or a character image
such that its edge emphasis means produces the edge emphasized signals
further in accordance with the image determination.
BRIEF DESCRIPTION OF THE INVENTION
The manner by which the above objects, features, and advantages are
attained will be fully apparent from the following detailed description
when it is considered in view of the drawings, wherein:
FIG. 1 is a block diagram showing an arrangement of an embodiment of an
area discriminating system for an image processing system according to the
present invention;
FIG. 2 shows an arrangement of IPS modules in the image processing system;
FIGS. 3(a) through 3(q) are explanatory diagrams for explaining the
respective modules of the IPS;
FIGS. 4(a) through 4(d) show a hardware configuration of the IPS;
FIGS. 5(a) through 5(d) show block diagrams for explaining an embodiment of
an edge processing system according to the present invention;
FIGS. 6(a) and 6(b) graphically show an arrangement of an edge processing
LUT;
FIGS. 7(a) and 7(b) are block diagrams showing hardware arrangements of a
non-linear filter section, which is constructed with LSIs;
FIGS. 8(a) through 8(g) are explanatory diagrams for explaining the
operation of the circuit shown in FIGS. 7(a) and 7(b);
FIG. 9 is a block diagram showing an arrangement of a first embodiment of a
sharpness improving system for an image processor according to the present
invention;
FIGS. 10(a) through 10(c) are diagrams useful in explaining the sharpness
improvement;
FIGS. 11 and 12 are block diagrams showing arrangements of other
embodiments of a sharpness improving system for an image processor
according to the present invention;
FIGS. 13(a) through 13(e) are block diagrams for explaining an embodiment
of an area discriminating method for an image processing system according
to the present invention;
FIGS. 14(a) through 14(c) are block diagrams of a large area discriminating
circuit based on the fixed block discrimination method;
FIGS. 15(a) and 15(b) are explanatory diagrams for explaining the variable
block discrimination method for making a distinction between the different
image areas on the basis of the edge interval;
FIG. 16 is a block diagram showing an arrangement of a large area
discrimination circuit using the variable block discrimination method;
FIG. 17 is a block diagram showing an arrangement of a digital color image
processing system;
FIG. 18 is a block diagram showing an arrangement of a conventional edge
processing circuit;
FIGS. 19(a) through 19(c) are block diagrams showing an arrangement of a
hue detect circuit;
FIGS. 20(a) through 20(c) are a graph and explanatory diagrams, useful in
explaining a character widening phenomenon; and
FIGS. 21(a) through 21(c) are explanatory diagrams for explaining edge
emphasis processing.
DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENT
An area discriminating system for an image processing system, as shown in
FIG. 1, includes an edge detect filter 1 as a high-pass filter for
detecting an edge quantity of a high frequency component. A hue detecting
circuit 2 detects hues, and an emphasis signal generating circuit 3
generates an emphasis signal for edge portions on the basis of the output
signal of the hue detect circuit 2 and the edge detect filter 1.
The image processing system controls the reproductivity of an image
according to the type of image by properly selecting parameters of the
emphasis signal generating circuit 3 to generate an edge emphasis signal,
and combining the edge emphasis signal with the output signal of the
smoothing filter 5 to generate a record signal. A large area
discriminating circuit 4 discriminates image areas for every block from
edge information o each picture element, and selects the parameters of the
emphasis signal generating circuit 3 according to an area.
With such an arrangement, even if the area determinations based on the edge
information of picture elements are not uniform, since the area
discrimination is performed for each block, the nonuniformity of the area
determination and mistaken determination can be corrected to improve the
accuracy of discrimination.
The large area discriminating circuit 4 calculates an average value of edge
quantities of picture elements whose optical density exceeds a
predetermined optical density threshold value of, and determines that an
image area is a character area or a halftone area depending on whether or
not the average value is above or below a threshold value of edge
quantity, or when optical density of all the picture elements is above or
below a preset density threshold value. The large area discriminating
circuit 4 calculates the ratio of the number of picture elements whose
density exceeds a preset density threshold value and the number or picture
elements whose edge quantity exceeds a preset edge threshold value, and
determines the image area to be a character area or a halftone area
depending on the calculated ratio. The large area discriminating circuit 4
determines the image area to be a character area or a half tone area
depending on the distance between the picture elements having
corresponding edge quantities exceeding a preset edge threshold value, and
further an average density value or a minimum density value.
Thus, the area discrimination is carried out on the basis of the picture
element density and the edge quantity. With such arrangements, the
character area featured in that where the background density is low, the
edge quantity is larger than the average value, the edge quantity is
large, and the like, and can readily be discriminated from the halftone
area having opposite features. In the embodiments to follow, a color
copying machine will be used as the image processing system. However, it
should be understood that printers, facsimile, and other types of image
processing apparatuses are involved in the image processing system.
I. IPS MODULES
FIG. 2 shows an arrangement of IPS modules in the image processing system
(IPS). In the color image recording apparatus, the IIT (image input
terminal) reads a color image on an original in the form of three primary
colors, B (blue), G (green) and R (red) by using a CCD image sensor, and
converts these to signals of primary toner colors of Y (yellow), M
(magenta), C (cyan), and K (black or tusche), and the IOT (image output
terminal) performs the exposure by the laser beam and development to
reproduce the original color image. In this case, the four separated toner
images of Y, M, C and K are used. A copy process (pitch) is performed one
time using the process color of Y. Subsequently, the copy processes will
be performed for the remaining process colors M, C and K. A total of four
copy cycles are executed. The four images consist of mesh points and are
superposed to reproduce a single full color image. Accordingly, when the
separated color signals of B, G and R are converted into toner signals of
Y, M, C and K, a designer encounters the problems how to adjust the color
balance, how to reproduce colors in accordance with the read
characteristic of the IIT and the output characteristic of the IOT, how to
adjust the balance of density and contrast, and how to adjust the emphasis
and blur of the edge, and how to adjust for Moire.
The IPS receives the separated color signals of B, G and R, processes the
signals to improve the reproducibility of colors, tone, and definition,
converts the toner signals of the developing process colors into on/off
signals, and outputs them to the IOT. As shown in FIG. 2, the IPS is made
up of an END (equivalent neutral density) conversion module 301, color
masking module 302, original size detect module 303, color conversion
module 304, UCR (under color removal)/black generating module 305, spatial
filter 306, TRC (tone production control) module 307,
reduction/enlargement processing module 308, screen generator 309, IOT
interface module 310, area image control module 311 including an area
generator and a switch matrix, and edit control module including an area
command memory 312, color palette video switch circuit 313, and font
buffer 314.
In the IPS, 8-bit data (256 gray levels) representing each or the separated
color signals B, G and R is applied to the END conversion module 301. The
module 301 converts the data into the toner signals of Y, M, C and K. A
process color toner signal X is selected and digitized. The digitized
signals are transferred, as the on/off data of the process color toner
signals, from the IOT interface module 310 to the IOT. Accordingly, in the
case of full color (4 colors), the prescan is executed to detect an
original size, an edit area, and other necessary information of the
original. Then, a first copy cycle is executed using Y as the toner signal
X of the process color. Then, a second copy cycle is executed using M for
the toner signal X. Subsequently, copy cycles will be executed for the
remaining process colors. A total of four copy cycles are repeated.
In the IIT, the color components of R, G and B of the image are read by
using the CCD sensor, with the size of one pixel being 16 dots/mm. The IIT
outputs the read signals as 24 bits of data (3 colors.times.8 bits; 256
gray levels). B, G and R filters are laid on the upper surface of the CCD
sensor with the density of 16 dots/mm and whose total length is 300 mm.
The CCD sensor scans 16 lines/mm at a process speed of 190.5 mm/sec.
Accordingly, the sensor produces the read data at the rate of about 15M
pixels/sec for each color. The IIT log converts the analog data of B, G,
and R pixels to obtain the density data from the reflectivity data, and
then digitizes the density data.
The respective modules will be described in detail. FIGS. 3(a) through 3(g)
are explanatory diagrams for explaining the respective modules of the IPS.
(A) END Conversion Module
The END conversion module 301 adjusts (converts) the optically read signal
of the color original obtained by the IIT into a gray balanced color
signal. The amounts of toner of each color are equal when the color is
gray. The toner amount of gray is used as a reference toner amount.
However, the separated color signals of B, G, and R produced from the IIT
when it reads the gray document, are not equal in value, because the
spectral characteristics of the light source and the color separation
filter are not ideal. These imbalanced color signals are balanced by using
a converting table (LUT: look up table) as shown in FIG. 3(a). This
balancing work by the LUT is the END conversion. When a gray original is
read, the LUT converts the B, G, and R color separated signals into
signals at the equal gradation in accordance with a level (black ->white)
of the gray image. The LUT depends on the characteristics of the IIT. 16
LUTs are used. Of those LUTs, all 16 tables are used for film projectors
including negative films, and 3 tables are used for copy, photograph, and
generation copy.
(B) Color Masking Module
The color masking module 302 converts the B, G, and R color signals into
signals indicative of toner amounts of Y, M, and C, respectively, through
a matrix operation. This conversion is applied to the signals after they
are subjected to gray balance adjustment by the END conversion.
In this instance, the conversion matrix for the color masking is a
3.times.3 matrix exclusively used for converting B, G, and R into Y, M,
and C. A matrix capable of dealing with BG, GR, RB, B.sup.2, G.sup.2 and
R.sup.2, in addition to B, G and R may also be used. Any other suitable
matrix may be used, if necessary. Two sets of matrices are used, one for
an ordinary color adjustment and the other for emphasis signal generation
in the monocolor mode.
Thus, when the video signal from the IIT is processed by the IPS, the gray
balance adjustment is first conducted. If it follows the color masking
process, the gray balance adjustment using the gray original must be made
allowing for the characteristics of the color masking. This makes the
conversion table more intricate.
(C) Original Size Detection Module
Originals to be copied may comprise not only standard size documents, but
also patched up documents and others. To select paper of a proper size
corresponding to the size of an original, it is necessary to detect the
size of the original. In case that the paper size is larger than the
original size, if the peripheral region of the original is masked, the
resultant copy will be excellent. For this reason, the original size
detection module 303 detects the original size at the time of scanning and
suppresses the platen color (edge suppress) at the time of scanning to
read the original image. Accordingly, a color, for example black, which is
clearly distinguished from the original is used for the platen color. The
upper limit value and lower limit value for the platen color
discrimination are set in a threshold register 3031, as shown in FIG.
3(b). At the time of a prescan, the signal is converted (gamma (r)
conversion) into a signal X representing the data approximate to the
reflectivity of the original (by using the spatial filter 306 to be
described in detail). The signal X is compared with the upper/lower limit
value set in the register 3031, by a comparator 3032. An edge detect
circuit 3034 detects the edge of the original, and stores the maximum and
minimum values of X and Y in the coordinates into a max./min. sorter 3035.
As shown in FIG. 3(d), when the original is slanted or its figure is not
rectangular, the maximum values and the minimum values (x1, x2, y1, y2) at
four points on the outline of the figure are detected and stored. At the
time of scanning to read the original, the comparator 3033 compares the Y,
M and C of the original with the upper/lower limit values in the register
3031. A platen color suppress circuit 3036 suppresses the pictorial
information outside the edge, viz., the read signal of the platen, to
effect the edge suppressing processing.
(D) Color Change Module
The color change module 304 enables a designated color in a specific area
of an original to be erased. As shown in FIG. 3(c), this module is made up
of a window comparator 3042, threshold register 3041, and color palette
3043. To effect color change, the upper/lower limit values of Y, M, and C
of the colors to be changed are set in the threshold register 3041. The
upper/lower limit values of Y, M, and C of the converted colors are set in
the color palette 3043. According to an area signal applied from the area
image control module, the NAND gate 3044 is controlled. When it is not a
color change area, the color signals of Y, M, and C of the original are
transferred intact from a selector 3045. When the color change area is
reached, and the color signals of Y, M, and C of the original are between
the upper limit values and the lower limit values as set in the threshold
register 3041, the selector 3045 is switched by the output signal of the
window comparator 3042 to send the converted color signals of Y, M, and C
that are set in the color palette 3043.
As for the designated color, by directly pointing an original by a
digitizer, 25 pixels of B, G, and R in the vicinity of the coordinates as
designated at the time of prescan are averaged and the designated color is
recognized on the basis of the average. By means of the averaging
operation, even in the case of an original with 150 lines, the designated
color can be recognized with a precision within 5 of color difference. To
the B, G and R density data, the designated coordinates are converted into
an address and the density data are read out of the IIT shading correction
circuit, with that address. In the address conversion, readjustment
corresponding to the registration adjustment is needed, as in the case of
the original size detection. In the prescan, the IIT operates in the
sample scan mode. The B, G, and R density data read out of the shading RAM
are subjected to a shading correction by a software, and averaged.
Further, the data are subjected to END correction and color masking, and
then are set in the window comparator 3042. The registered colors are
selected from 1670 colors, and up to eight colors can be simultaneously
registered. The reference color prepared include a total of 14 colors, Y,
M, C, G, B, and R, colors between these colors, and K and W.
(E) UCR/Black Generation Module
When the color signals of Y, M, and C have equal quantities, gray is
produced. Theoretically, the same color can be obtained by replacing the
colors of Y, M, and C of equal quantities with black. In this case,
however, the color is impure and hence the reproduced color is not fresh.
To cope with this problem, the UCR/black generation module 305 generates a
proper amount of K to prevent such a color impurity, and equally reduces
the toner colors Y, M, and C in accordance with the amount of the
generated K (this process is called an under color removal (UCR)). More
specifically, the maximum and the minimum values of the toner colors Y, M,
and C are detected. A value of K is generated by a conversion table in
accordance with the difference between the maximum value and the minimum
value. Further, the toner colors Y, M, and C are UCR processed in
accordance with the generated K.
As shown in FIG. 3(e), in the case of a color closer to gray, the
difference between the maximum and the minimum values is small.
Accordingly, the minimum value or its near value of each color Y, M, and C
is removed for generating the color K. When the difference is large, the
removal quantities of the colors Y, M, and C are set below the minimum
values of them, thereby to reduce the quantity of the generated K. In this
way, the mixing of tusche into the pure color and the hue degradation of a
low gradation, high hue color can be prevented.
FIG. 3(f) shows a specific circuit arrangement of the UCR/black generation
module, a max./min. value detector 3051 detects the maximum and the
minimum values of the process colors Y, M, and C. A calculating circuit
3053 calculates the difference between the maximum and the minimum values
of each color.
A conversion table 3054 and another calculating circuit 3055 cooperate to
generate the black value K. The conversion table 3054 adjusts the value of
K. When the difference between the maximum and the minimum values is
small, the output signal of the conversion table is zero. Accordingly, the
calculating circuit 3055 produces the minimum value as intact in the form
of the value of K. When the difference is large, the output value of the
conversion table 3054 is not zero, the calculating circuit 3055 subtracts
the difference from the minimum value and produces the result of the
subtraction as the value of K.
A conversion table 3056 provides the values to be removed from the colors
Y, M, and C in accordance with the K value. In cooperation with the
conversion table 3056, an additional calculating circuit 3059 subtracts
the values as defined by the K value from the process colors Y, M, and C.
The AND gates 3057 and 3058 operate for the signal K, and the signals of
Y, M, and C after UCR processing in accordance with the signals in the
monocolor mode and the full color mode. The selectors 3052 and 3050 are
used for selecting any of the toner signals Y, M, C, and K by the process
color signals. A color is thus reproduced by using the mesh points of Y,
M, and C. Accordingly, the curves and tables that are empirically formed
are used for the removal of Y, M, and C and for determining the generation
ratio of K.
(F) Spatial Filter Module
In the color image recording apparatus incorporating the present invention,
the IIT reads an image of an original while the original image is being
scanned by the CCD. When the data is used as intact, the resultant data
will in effect be faded data. The mesh points are used for image
reproduction. Accordingly, Moire occurs between the mesh point period of
the printed matter and the sampling period of 16 dots/mm. The same
phenomenon occurs between the mesh point period generated by the machine
and that of the original. The spatial filter module 306 is provided to
remove the above fading and the Moire phenomenon. For the Moire removal, a
low-pass filter and for edge emphasis, a high-pass filter are used.
In the spatial filter module 306, as shown in FIG. 3(g), a selector 3003
selects one of the input signals Y, M, C, Min, and Max-Min. A conversion
table 3004 converts it into data signals approximately indicative of the
reflectivity. Use of this type of data makes it easy to pick up the edge
data. In this instance, the selected color signal is Y. A threshold
register 3001, 40 bit digitizer 3002, and decoder 3005 separate the color
signals Y, M, C, Min, and Max-Min into eight colors, Y, M, C, K, B, G, R,
and W (white), for each pixel. A decoder 3005 recognizes the hue in
accordance with | | |
